Diabetic neuropathy, once well established, is poorly reversible in human trials. Sensitive diagnostic measures are necessary to identify patients with very early axonal injury who may be better candidates for intervention or even preventative therapy. The Cutaneous Measures of Diabetic Neuropathy (CMND) study extensively characterizes peripheral nerve function in 221 diabetic subjects, with longitudinal follow-up. CMND shows that early diabetic neuropathy is characterized by progressive small fiber loss, and identifies skin biopsy with intraepidermal nerve fiber density (IENFD) as a sensitive, quantitative surrogate measure of neuropathy severity. IENFD closely correlates with clinical and electrodiagnostic measures of neuropathy, and declines significantly over two years in diabetic subjects, including those without neuropathy symptoms. We have shown that metabolic improvement achieved through individualized diet and exercise counseling for subjects with early neuropathy associated with impaired glucose tolerance significantly improves measures of small fiber neuropathy, including pain and IENFD. We hypothesize that 1) early IENFD decline predicts future development of clinically meaningful diabetic neuropathy, and 2) correction of IENFD decline with aggressive lifestyle intervention based on the impaired glucose tolerance neuropathy studies will successfully result in weight loss, improved metabolic parameters, and reduced risk of diabetic neuropathy development or slowing of its progression. Aim 1 will follow CMND diabetic subjects for an additional four years to determine the magnitude of early decline in IENFD that predicts future development of symptomatic distal polyneuropathy. Aim 2 will randomize 175 additional diabetic subjects without neuropathy symptoms to determine if two years of aggressive diet and exercise counseling reduces IENFD progression, or incidence of symptomatic neuropathy. Both Aims will examine if metabolic syndrome, obesity or specific baseline metabolic parameters, insulin resistance (HOMA-IR) or adipokine levels predict rate of IENFD decline or risk of future neuropathy development. PUBLIC HEALTH RELEVANCE Injury to the very longest peripheral nerves due to diabetes (diabetic neuropathy) is very common. The earlier diabetic neuropathy can be identified, the more effective treatment is likely to be. This proposal will test a method to count nerves in skin to see if it predicts future diabetic neuropathy, and tests the effect of diet and exercise on preventing nerve injury.